International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume:07 Issue: 09 | Sep 2020 www.irjet.net e-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1013
MACHINE LEARNING APPROACH BASED HOME ENERGY MANAGEMENT
PREDICTION SYSTEM
P.Naveenkumar
1
, G.Rajeswari
2
1
Assistant Professor, Department of Electrical and Electronics Engineering, Knowledge Institute of Technology,
Tamil Nadu, India
2
PG Scholar, Department of Electrical and Electronics Engineering, Knowledge Institute of Technology,
Tamil Nadu, India
----------------------------------------------------------------------------------------------------------------------------------
Abstract - Energy crisis is one among the prime
challenges being faced by many of the countries within the
world today. For an enormous extent of the demand of
energy in industrial development has increased
tremendously. A lot of techniques are suggested like an
Energy monitoring and prediction system which is an
efficient technique to watch the devices present inside a
house or industries and provide notification about their
abnormal behavior. In this paper, we have focused on
predicting electric energy use of home appliances in a low
energy consumption house. Electric energy demands are
changed in weekdays and weekend days due to the staying
time of home residents. In this project the implementation
of an advanced Internet of Things (IoT) based system for
intelligent energy management in Industries and the home
usage. The users can view their status threw the IOT based
android application and the webserver.
Key Words: Energy Meter, Intelligent Prediction
Model, Internet of Things. Energy Management
1. INTRODUCTION
Internet of Things and Everything (IoT/E) is a relatively
new concept that explains the aggregation of objects,
systems, processes etc, via an IPv6 backbone into the
web environment. In the existing power utility setup, an
IoT framework is often to improve the issues faced by
energy consumers on day to day. For instance,
consumers are presented with consumed energy details
just one occasion a month with zero control on the
deployed smart meter (if in existence). The way that is
too long for the consumer to watch the instantaneous
changes within the power usage. So the length of the
time is also getting increase. In addition, utility bills can
be convoluted on how they present usage information,
but a consumer may not be able to decipher changes in
their power usage from the last bill. In most cases, if any
consumer fails to pay the bill, the utility field marketers
go to their houses to disconnect the facility supply. To
display the total real time usage of power and amount of
electricity, the conventional energy meters are presently
installed in households.
1.1 Need for Demand
There is no thanks to access the small print on
what the day’s, week’s or month’s consumption was on
these meters. Often, these meters are placed in an
inconvenient location which makes regular viewing
somewhat difficult. Sadly, ubiquitous access to
consumption rate and energy profile by consumers is
impracticable. By deploying a sensing/monitoring
mechanism in real time (using IoT), the precise power
consumed are often ascertained while deriving informed
decisions on the way to manage the connected loads. A
system which will give users an estimate of what
proportion energy is getting used, that has been used, or
which may be consumed will allow them to regulate
their habits and lower the costs [4]. This is a sort of
Demand Side Management during a previous study, i.e.,
R-SGEMS.
1.2 Meter Reading Techniques
In the conventional electro-mechanical and
computerized metering framework, electric vitality is
reviewed by individual and regularly they arranged the
bill through suspicion dependent on his history of power
utilization. Possibly the buyer has not used the
comparable measure of power in the present month as in
the earlier months for reasons, for example, holidaying
somewhere else or being in the office for long time, and
so on. This strategy for charging is additionally not
appropriate for the power supply organization since it
gives a wrong record of the general power utilization in
the buyer's region and may at last outcome in blunders
in future arranging by the organization.
Over the previous years, metering gadgets have
experienced numerous enhancements and clothed to be
more convoluted with more highlights and capacities.
Electro mechanical Meter has next to no exactness and
absence of configurability. There are such a significant
number of issues require service organizations to